Wishful Thinking: A Critical Analysis of Homeopathy Part II

Homeopathic 1 gram globules

This is a continuation of my post on homeopathy in which I talked about the history of homeopathy and how it is scientifically tested.  I ended the post wondering why there was so much public acceptance for such an ineffective product.  In this post, I will discuss the various cognitive biases that allow well-intentioned individuals to falsely believe that homeopathic interventions have positive effects.

What about the Public?

If homeopathy has such a poor track record with controlled studies and with generating consistent and plausible mechanisms of action, why is it so popular?  Well, when it was invented, it probably caused fewer deaths and complications than the prevailing medical knowledge at the time.  Eating ice-cream would beat blood-letting as a cure for pretty much everything (except perhaps for toomuchbloodosis).  Hence, homeopathy might actually have saved more lives than conventional medicine when it was invented, simply due to the fact that it didn’t do anything (harmful or helpful).  On the other hand, nowadays we have a plethora of modern treatments for a vast number of diseases based on our refined understanding of human biology, physiology, immunology and a string of other “ogy”s that require many years of education and effort to master.  Given this fact, why is homeopathy popular today?  The rich, the poor, the educated, the illiterate, the powerful and the oppressed – all sorts of people have tried homeopathy or know someone who has.  Strangely enough, reputable doctors sometimes stand behind homeopathy and the British government (through the NHS) used to fund homeopathic interventions and treatments using tax-payer funds (until recent backlash).  What would prompt consumers to take up a treatment that is so implausible?  The answer lies in the way our brains work. The human brain is far from perfect when it comes to evaluating events, especially when they involve complex statistical distributions.  We suffer from a number of cognitive biases that are a product of our evolutionary history such the confirmation bias, framing effect, illusion of control and neglect of probability.  I discussed some of these biases here, in the context of rational economic action.  In this case, the major effects that contributes to a favorable view towards homeopathy are the placebo effect, confirmation bias, and regression to the mean.

The Placebo Effect

The placebo effect is a complicated amalgam of various neurological and psychological effects that allow the subjects of impotent interventions to report a subjective improvement of the allegedly treated symptoms/conditions.  A subject given a sugar pill and told that it is a muscle relaxant tends to report that their muscles do feel more relaxed.  This is a mental trick – the brain is responsible for constructing and synthesizing our reality and it can easily be manipulated into changing our perceptions.  It is widely known and good medical studies use a placebo control group to account for its effects when testing a new treatment or drug.  Homeopathy tends to produce a placebo effect in its users because users of simple symptoms tend to subconsciously believe that the treatment that they are getting is effective and is helping them.  A good demonstration of this effect is in the various effects grouped under alcohol expectancy effects.  Human participants, when given a drink and told that it is alcoholic, tend to report feeling tipsy or act under the influence even if not actually given an alcoholic drink.  The effect is nowhere near the effect seen and reported when subjects are actually given alcohol, but it behooves us to be aware of the role our brain and perception play into interpreting and reporting events.

Confirmation Bias

Confirmation bias is the tendency of individuals to tend to selectively interpret or manipulate their environment to conform to their preconceived notions.  If an interviewer thinks that a candidate is not worthy of a job offer before they conduct an interview, they are more likely to try to subconsciously try to confirm this by finding data that makes the candidate look worse.  If they instead believed a candidate to be worthy before an interview, they are more likely to ignore their shortcomings and emphasize their strengths.  Hence, interviewers need to be extremely careful and aware of this effect during interviews and try to avoid it.  People using medicine act like interviewers interviewing new candidates.  They tend to want to have their preconceived notions confirmed.  They do not have the time or energy to perform complex statistical analysis on their lives and their medical issues.  After all, who wants to take a medicine and think “I wonder if I will get sicker” or “I wonder if this is a waste”?  Most people are trying to get better, not worse.  Hence, when presented a recipe that they hope will work, humans in general will try to confirm their preexistent notions.  They are not being malicious or facetious.  They are simply performing a task (one of discerning the effectiveness of a remedy) using primitive tools – their brains alone.

Regression to the Mean

Regression to the Mean - from Norwegian University of Science and Technology

Regression to the mean is a mathematical illusion or artifact when dealing with statistical distributions.  When dealing with measurements that are based on an underlying statistical distribution, an extreme measurement tends to be followed by a measurement that is closer to the mean.  For example, consider a video game (Wii bowling…) that can be repeatedly played and in which each round is identical.  You may quickly notice that whenever you get a round in which you happened to have performed really, really well, you tend to perform much worse on the next round.  Rounds during which you perform really, really badly tend to be followed by rounds during which you perform better than before.  This is because unlikely events are, by definition, unlikely and tend to be followed by more likely events.  Suppose that you notice this trend, and after each round you did badly on, you decide to perform a song and dance and praise a good luck charm.  If you then conclude that the fact that you do better on the next round is due to your song and dance, you have committed a fallacy.  Your game was highly likely to get better anyway, due to the laws of statistics, and the song and dance did not significantly contribute to this at all.  This effect comes mostly into play when people are trying to recall or collate stats about their homeopathy use.  People tend to get medicine when they are sick.  It is also true that for a lot of simple illnesses, people simply get better over time.  If they happen to be taking a homeopathic remedy during this time, they may associate their natural healing power with a (non-existent) power of the homeopathic remedy.  They should just be giving themselves the credit.

Final Words

I hope that through reading this, you have an increased understanding of the extraordinary claims that homeopathy makes and the lack of evidence for most of these claims.  I hope that you have taken away a better understanding on what it takes to evaluate medical hypotheses and how the human mind can be easily tricked into believing a number of fallacious ideas.  If you do use homeopathy, I sincerely hope to have convinced you to, at the very least, question the efficacy of the medication you are currently taking and think about whether it is worth the costs.  If you know people you love and care who are currently using homeopathy to cure life threatening diseases, please talk to them!  Help them understand what it is that they are doing and get proven clinical treatments (if they exist).  Homeopathy may appear to be harmless (after all, it’s just water, alcohol or sugar) but a hidden toll is taken on people who really do need proven medicine to cure life threatening illnesses but are instead spending money on ineffective treatments with a false sense of hope.